Reference & Advocacy Cultivation Finder
Five references for the new launch — in 90 minutes, ranked by health, with hooks.
Sample output for this workflow will appear here once it is captured.
Run the workflow in Claude, ChatGPT, or Phoenix Playground using the buttons below to see real output.
Overview
Marketing needs 5 references in your fastest-growing vertical and you've got 90 minutes. Pass in your customer list + target segment, the workflow runs each customer through the `csm-health-scoring` composite, ranks them by health, and surfaces the top candidates with conversation hooks: how long they've been healthy, which milestone they hit, what their public posture is. Hard cap at 5 customers per run — fan-out beyond that exceeds the tool-call budget.
Use cases
Marketing asks for 5 FinServ references — you ship in 90 minutes
Lin runs this with `target_segment = FinServ enterprise` and her 5 healthiest FinServ customers as `customer_domains`. The output ranks the top 3 with hooks: 'Healthy for 7 quarters, hit Q2 adoption milestone, CISO posted publicly about the category last month'. She emails marketing in an hour with the list and the hooks.
Health composite is consistent with the churn workflow
The same `csm-health-scoring` rubric that grades accounts in the scorecard is what surfaces reference candidates here. Customers who are 'healthy enough to expand' are the same ones 'healthy enough to ask for a reference'. No drift between workflows.
View workflow prompt
# Reference & Advocacy Cultivation Finder
## Parameters
- `{{customer_domains}}` *(required)* — Comma-separated existing customer domains. Hard cap 5 per invocation. HG tools cannot identify customers from the open market — input list IS the candidate pool.. Example: `customer1.com,customer2.com,customer3.com`
- `{{target_segment}}` *(required)* — Industry / size / geography filter for which customers in {{customer_domains}} qualify. Example: `FinServ enterprise`
## Purpose
Marketing needs references in {{target_segment}}. Pass in {{customer_domains}} (your CRM-supplied existing-customer list, hard cap 5 per run). The workflow filters by segment, ranks survivors by the `csm-health-scoring` composite, and surfaces the top 3 with conversation hooks. HG MCP cannot identify existing customers from the open market — that's why {{customer_domains}} is required.
## Process
1. **Segment filter** — `company_firmographic` for each domain in {{customer_domains}}. Filter to those matching {{target_segment}} (industry / size / geography).
2. **Health composite** — for each surviving domain, run the `csm-health-scoring` rubric (adoption + sponsor + intent + sentiment). `company_install_time_series`, `company_fai`, and `company_intent` provide the components. *company_contracts* is NOT used for renewal weight (outsourcing-only); renewal-cycle weight is "unknown" unless CRM supplies the date — which this workflow does not require.
3. **Hooks** — for the top 3 surviving domains by composite, identify 1-2 advocacy hooks: longest period at High, hit a specific public milestone (e.g., an IPO), CEO posted publicly about the category, intent rise on adjacent topics (signals expansion appetite).
4. **Rank** — output the 3 candidates ordered by composite score; tie-break on hook strength.
## Output Format
- `# 🤝 Reference Candidates for {{target_segment}}`
- `## Cohort filtering` — N of {{customer_domains}} matched {{target_segment}}
- `## Top 3 candidates` — ranked table: rank | domain | composite | top hook | confidence
- `## Outreach hooks` — 1-2 sentences per candidate naming what to lead with
## Quality Checklist
- {{customer_domains}} hard-capped at 5 — instruct user to re-run for larger lists (each customer takes ~3 tool calls)
- Composite cites every component
- No fabricated hooks — every hook traces to a specific source (firmographic event, intent rise, FAI stability)
- *company_contracts* not used in this workflow — renewal-cycle weight is "unknown" by design (no CRM date input)
- Cap total tool calls at ~15